The second edition of this textbook (henceforth SIEG2) was published twenty years after the publication of the Cambridge Grammar of the English Language (CGEL) on which it is based, and seventeen years after the first edition (SIEG1). The latter has acted as an introduction to the larger work for many generations of students. I use the book on a Master's course in English linguistics. Quite a few students find it challenging, not only because many of its analyses are different from what they have been taught on previous undergraduate degrees, but also because the book's account of certain areas of English grammar can be quite complex.
How is the second edition different? SIEG2 is published in a new, much more attractive format and layout. It has an additional chapter on adjuncts, and some of the material in SIEG1 has been moved online, specifically the chapter on morphology and the glossary. The text of the other chapters is largely the same, but has been updated. There are quite a few further notable changes, some of which I will discuss below, concentrating on exposition, terminology and the online support materials.
Exposition
The text of this edition has become more accessible, and here and there tree diagrams are added for clarity. As an example, consider the tree diagrams for the NPs the guy who fainted and a young woman, shown in (1a) and (1b) (p. 113):
(1)
As in CGEL, the trees in SIEG2 show both form and function labels. However, unlike in CGEL, they make explicit that a single word on its own can constitute a phrase. In Aarts (Reference Aarts2004: 370) I criticised CGEL for representing a single determinative as just ‘D’ in tree diagrams, and asked why it is not a determinative phrase (DP) as well. That was unjustified, because I didn't spot an easy-to-miss comment on page 329 in CGEL: ‘We simplify the tree diagrams by omitting the higher-level constituents if they consist of just a head element.’ The fully explicit trees in SIEG2 avoid this potential misunderstanding. However, confusion may still occur for some readers over the label ‘DP’ which is also used in generative frameworks in an entirely different way; ‘DetvP’ would perhaps have been better.
A very welcome new chapter on adjuncts (which had been circulating as a pdf for quite a few years) has been added. The chapters in SIEG2 now align with CGEL.
Terminology
In this section I will discuss some issues concerning the use of terminology in SIEG2.
Head
In SIEG1 the concept of ‘head’ is defined as follows: ‘The head of a phrase is, roughly, the most important element in the phrase, the one that defines what sort of phrase it is’ (p. 13). This is replaced in SIEG2 by the following: ‘A phrase in our terms is a constituent … with a word functioning as head and some number (zero or more) of dependents’ (p. 23, emphasis in original). This is a more useful definition, but note that neither definition caters for the VP disturbs me in the tree diagram in (2) which functions as the head of the VP that immediately dominates it, which in turn functions as predicate.
(2)
Here the function of head is realised by a phrase, not by a word. (The same point applies to other phrase structures, for example to noun phrases, where ‘Nom’ functions as head; see the tree diagram of a young woman in (1b).) Incidentally, notice that in (2) the extraposed subject clause that he was acquitted is treated as a complement, positioned within the higher Predicate VP. The reader may wonder what this clause is a complement of, given that complements are defined in an earlier chapter as ‘an integral and sometimes obligatory part of a phrase that the specific lexical head permits or requires’ (p. 23). The structural configuration in the tree in (2) (VP-adjunction) does not make clear which lexical head licenses this complement. One might say that the extraposed clause is ‘indirectly licensed’ by the verb disturb, but this would need to be made explicit, as the tree does not represent the usual head-complement configuration. CGEL (p. 1403) states that ‘[a]n extraposed subject, like a displaced subject, is not a kind of subject, but an element that is related to a dummy subject’, and it refers to ‘the extraposed subject position’ as being ‘at the end of the matrix clause’, without further specification. This is unhelpful, because it does not tell us exactly where the extraposed clause is placed. As the tree in (2) shows, in SIEG2 we have more detail than in CGEL, but I would have liked to see some justification for regarding the extraposed clause as a complement, and for positioning it within the higher VP. The latter could easily have been achieved with a standard VP constituency test: I said that it disturbs me that he was acquitted and [VP disturb me that he was acquitted] it did.
In SIEG1 the coordination Jennifer is a great teacher and her students really respect her is represented as in (3):
(3)
In SIEG2 the structure has been revised to look like (4) (p. 342):
(4)
The earlier structure was problematic (not least because two constituents carry the identical label ‘Coordinate2’), but the newly proposed structure is also not without problems, because the text reads: ‘The coordinates … have equal syntactic status: each makes the same sort of contribution to the whole thing.’ But this is not what the tree diagram shows, since the clause and her students really respect her has a different structure: it is introduced by a coordinator which grammatically functions as marker. Notice also the label ‘coordination’ at the top of the diagram, which was not explained in SIEG1. In the new edition the authors write: ‘We're essentially using the term coordination as the name of a syntactic category to which all coordinations belong – a category that is neither lexical nor phrasal’ (p. 342). But this does not help much, because the exact syntactic nature of this category is still left unexplained. We have one further kind of unit in CGEL and both editions of SIEG that is neither lexical nor phrasal, and this is ‘Nom’ (N-bar) inside noun phrases, but clearly ‘coordination’ is not on a par with Nom.
Finite(ness)
The discussion of the notion of finiteness has been moved from chapter 3 in SIEG1 (‘Verbs, tense, aspect and mood’) to chapter 14 in SIEG2 (‘Non-finite clauses’). This means that it is not until readers get to page 311 of the book that this important grammatical concept is explained. In line with recent work in linguistics, CGEL (p. 88) regards finiteness as a property of clauses, rather than of verbs, because the traditional definition ‘doesn't work at all for present-day English’ (SIEG2: 311; see also e.g. Nikolaeva Reference Nikolaeva and Nikolaeva2007). In CGEL and SIEG1/2 finiteness is a property of clauses that contain a primary verb form, and of imperative and subjunctive clauses. Dropping the traditional definition of ‘finite clause’ as ‘a clause headed by a finite verb’ entails that CGEL and SIEG1/2 allow for tensed non-finite clauses, as in I was happy to have met him, where the underlined clause contains a secondary perfect tense. CGEL and SIEG1/2 also allow for finite clauses to lack a tensed verb, as in subjunctive and imperative clauses. While all this makes sense to those who are well versed in English grammar, it is hard for students to understand, especially if they have previously been taught the traditional definition of finiteness.
Particle
The label ‘particle’ is used in almost all grammars as a word class label, so students can be forgiven for misunderstanding this notion in SIEG2, where it is used as a grammatical function label, so that in the sentence He switched the light off, the verb switch takes two internal complements, namely the NP the light, which functions as direct object, and the prepositional phrase off (headed by an intransitive preposition), which functions as particle. The discussion in SIEG2 is an improvement on SIEG1, because the fact that ‘particle’ is a function label is more clearly signposted, but I believe that this will still cause a great deal of confusion, especially when we read that ‘[d]erivatively, we can call a word a particle if it has the potential to function as a particle complement’ (p. 199, emphasis in original). I'm really not sure what the authors have in mind here. None of this is helped by the fact that the chapter in which the discussion of ‘particle’ occurs is entitled ‘Prepositions and particles’, which mixes form and function labels.
Catenative verb
Catenative verbs have played a role in English linguistics at least since Palmer (Reference Palmer1965/1987) and Huddleston's earlier work (Reference Huddleston1984). These verbs also play a major role in CGEL and SIEG1. In SIEG2, the label ‘catenative verb’ is dropped. I have mixed feelings about this. On the one hand, I argued in Aarts (Reference Aarts2004: 371) that it is superfluous, and can be confusing for students. I stand by that view, but removing the concept from this textbook will pose problems for students who wish to study this area of grammar in greater detail by consulting CGEL. They will find that the CGEL treatment of verbs occurring in this construction uses very different terminology from SIEG2 (apart from ‘catenative verb’, CGEL also has the labels ‘catenative complement’, ‘catenative-auxiliary analysis’ and ‘catenative construction’ – the latter occurs in various guises: simple/complex/oblique/for).Footnote 1
Transparent verb
This label is used in the chapter on non-finite clauses (p. 325) to designate raising verbs, as in Al appeared to like Ed. The authors explain that in this sentence ‘[i]t's as if the intervening verb were merely some kind of modifier – as if it were transparent to the subject–verb semantic relation. The meaning … is very close to that of Al apparently liked Ed, where the adverb apparently modifies the like VP’ (p. 325). I think that quite a few students will be puzzled by this, not least because it does not follow from the fact that we can paraphrase this sentence in this way that the ‘transparent verb’ is syntactically like a modifier. Moreover, with a nod to Occam, the new terminology is unwarranted, because SIEG2 retains the label ‘raised subject’, so this kind of verb might as well have kept the label ‘raising verb’, as in CGEL and SIEG1. And here again, as with ‘catenative verb’, CGEL does not use the label ‘transparent verb’, so it will again confuse students who wish to ‘level up’ to the larger work.
Online support materials
SIEG2 offers students and instructors copious online support materials (cambridge.org/SIEG2). For students there are pdfs with chapter summaries, supplemental tree diagrams for chapters 3–5, a glossary, lists of determinatives, prepositions and grammatical functions, a morphology appendix (formerly a chapter in SIEG1), and a document with further reading suggestions. For instructors (who need to apply for access), there are solutions to all the exercises, a ‘test bank’ for all the chapters in the form of multiple-choice exercises, and files for all figures in the book, made available in jpeg and ppt formats.
These materials are immensely valuable, though to my mind some of them should have remained inside the book, e.g. the glossary. It is irritating for readers to have to go online to check the meaning of a particular concept, when they could have just turned to the end of the book.
Although the ‘Further Reading’ document is very useful indeed, it would have been even more valuable had it been more comprehensive. As things stand, what we get is ‘a miscellaneous assembly of things’, but ‘[s]pace limitations made it quite impossible for us to provide in print the kind of full bibliographical referencing that our book would have in an ideal world, and the sheer bulk of the enormous linguistics literature pertaining to English makes it impossible for us to be complete here either’ (p. 1). But this is odd, given that the document is posted online, where there aren't any space limitations.
The document with fifty-two ‘supplemental tree diagrams’ clarifies the analyses adopted in the book. This will be enormously helpful to students and instructors too. Intriguingly, these tree diagrams also offer an exegesis of many of the analyses in CGEL, which has only forty tree diagrams, so that in many cases SIEG2 is more explicit in visualising the analyses of particular constructions than the larger grammar. Maybe it's also time for a second edition of CGEL.
Conclusion
In many ways this second edition is a much-needed improvement on the first: its writing is more accessible, there are more exercises, and the analyses proposed have been made much more explicit by using more tree diagrams, both in the text and in the online pdf.
Unfortunately, not all the changes are to be welcomed. This is true especially with regard to some of the new terminology used in the book which is not in sync with CGEL, while some other terminology has been dropped, making ‘levelling up’ to CGEL harder. The book is designed for a one-semester course, but it will be challenging for instructors to cover all the chapters within that time period. Two semesters are more appropriate for teaching the material effectively and at a more manageable pace.
